Survival Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mô tả về dataset Telco_Churn\n",
    "\n",
    "Các dữ liệu Telco_Churn là từ hồ sơ thông tin khách hàng của một công ty viễn thông. Mục tiêu của chúng tôi là dự đoán những khách hàng có thể ngừng sử dụng dịch vụ của họ.\n",
    "\n",
    "Mục tiêu của bài toán là dự đoán Churn. Với Churn trong trường hợp này là Customer Churn, hiểu đơn giản là khi một khách hàng quyết định ngừng sử dụng một dịch vụ hay kết thúc hợp đồng (không muốn gia hạn) với một công ty nào đó. Churn thường được sử dụng trong các công ty kinh doanh với mô hình thanh toán theo kỳ (subscriber-based service model).\n",
    "\n",
    "Ví dụ như khi các bạn quyết định ngừng sử dụng mạng di động X để chuyển qua dùng mạng của hãng Y. Lúc này các bạn sẽ được coi là churn của công ty X và là khách hàng mới của công ty Y. *Đọc thêm về Churn và Churn Rate* https://data-fun.com/cach-tinh-ty-le-churn-rate-la-gi/\n",
    "\n",
    "Thông thường bài toán này được giới Data Science áp dụng các mô hình phân loại như DecisionTree, RandomForest, Logistic Regression,... Tuy nhiên, bài viết của mình xin được phép giới thiệu một hướng phân tích khác. Đó là áp dụng Ước tính Kaplan-Meier trong phân tích vòng đời của khách hàng. \n",
    "\n",
    "-------------------------------------------------------------\n",
    "Kiến thức người viết hạn hẹp, cũng như bài viết chỉ mang tính chia sẻ học thuật. Mong được anh em góp ý thêm\n",
    "\n",
    "#### Lê Đặng Trung Hiếu\n",
    "*Phòng ABS, CTCP Chứng Khoán Bản Việt (VCSC)*\n",
    "\n",
    "Tầng 8, Sailing Tower, 111A Pasteur, Quận 1, TpHCM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import hieu_viet_code_ne"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Quan sát qua  bảng dữ liệu Telco_Churn\n",
    "\n",
    "Mọi người cũng có thể tải về data này tại: https://www.kaggle.com/blastchar/telco-customer-churn\n",
    "Hoặc sử dụng file raw trên github của mình https://raw.githubusercontent.com/cafechungkhoan/chu_gia/master/telco_churn.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>customerID</th>\n",
       "      <th>gender</th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>Partner</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>PhoneService</th>\n",
       "      <th>MultipleLines</th>\n",
       "      <th>InternetService</th>\n",
       "      <th>OnlineSecurity</th>\n",
       "      <th>...</th>\n",
       "      <th>DeviceProtection</th>\n",
       "      <th>TechSupport</th>\n",
       "      <th>StreamingTV</th>\n",
       "      <th>StreamingMovies</th>\n",
       "      <th>Contract</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7590-VHVEG</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>No</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>30</td>\n",
       "      <td>30</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5575-GNVDE</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>34</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>57</td>\n",
       "      <td>1890</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3668-QPYBK</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>2</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>54</td>\n",
       "      <td>108</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7795-CFOCW</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>45</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>No</td>\n",
       "      <td>Bank transfer (automatic)</td>\n",
       "      <td>42</td>\n",
       "      <td>1841</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9237-HQITU</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>2</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Fiber optic</td>\n",
       "      <td>No</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>71</td>\n",
       "      <td>152</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   customerID  gender  SeniorCitizen Partner Dependents  tenure PhoneService  \\\n",
       "0  7590-VHVEG  Female              0     Yes         No       1           No   \n",
       "1  5575-GNVDE    Male              0      No         No      34          Yes   \n",
       "2  3668-QPYBK    Male              0      No         No       2          Yes   \n",
       "3  7795-CFOCW    Male              0      No         No      45           No   \n",
       "4  9237-HQITU  Female              0      No         No       2          Yes   \n",
       "\n",
       "      MultipleLines InternetService OnlineSecurity  ... DeviceProtection  \\\n",
       "0  No phone service             DSL             No  ...               No   \n",
       "1                No             DSL            Yes  ...              Yes   \n",
       "2                No             DSL            Yes  ...               No   \n",
       "3  No phone service             DSL            Yes  ...              Yes   \n",
       "4                No     Fiber optic             No  ...               No   \n",
       "\n",
       "  TechSupport StreamingTV StreamingMovies        Contract PaperlessBilling  \\\n",
       "0          No          No              No  Month-to-month              Yes   \n",
       "1          No          No              No        One year               No   \n",
       "2          No          No              No  Month-to-month              Yes   \n",
       "3         Yes          No              No        One year               No   \n",
       "4          No          No              No  Month-to-month              Yes   \n",
       "\n",
       "               PaymentMethod MonthlyCharges  TotalCharges  Churn  \n",
       "0           Electronic check             30            30     No  \n",
       "1               Mailed check             57          1890     No  \n",
       "2               Mailed check             54           108    Yes  \n",
       "3  Bank transfer (automatic)             42          1841     No  \n",
       "4           Electronic check             71           152    Yes  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eda = pd.read_csv(r\"https://raw.githubusercontent.com/cafechungkhoan/chu_gia/master/telco_churn.csv\")\n",
    "eda = pd.DataFrame(eda)\n",
    "eda.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ước tính Kaplan-Meier\n",
    "\n",
    "Ước tính Kaplan-Meier (thường sử dụng trong y học) là một trong những phương pháp được đánh giá là tốt nhất được sử dụng để đo tỷ lệ đối tượng sống trong một khoảng thời gian nhất định sau khi điều trị. Trong các thử nghiệm lâm sàng hoặc thử nghiệm cộng đồng, hiệu quả của một can thiệp y tế nào đó được đánh giá bằng cách đo số lượng đối tượng sống sót hoặc được cứu sau can thiệp đó trong một khoảng thời gian. \n",
    "\n",
    "Nền tảng người tạo ra ước lượng này: Paul Meier sinh ngày 24/7/1924 tại Newark (New Jersey), tốt nghiệp cử nhân toán từ trường Oberlin vào năm 1945, tiến sĩ thống kê học từ Đại học Princeton năm 1951 dưới sự hướng dẫn của John Tukey.\n",
    "\n",
    "Main idea của phương pháp: Ước tính Kaplan-Meier là cách đơn giản nhất để tính toán sự sống sót theo thời gian bất chấp tất cả những khó khăn liên quan đến các đối tượng hoặc tình huống. Survival Curve tính toán xác suất xảy ra sự kiện tại một thời điểm nhất định và nhân xác suất kế tiếp này với xác suất được tính toán trước đó để có được ước tính cuối cùng. \n",
    "\n",
    "*Bảng research về survival analysis: Ước tính Kaplan-Meier* https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3059453/\n",
    "\n",
    "*bài giảng của Đại học Princeton: Survival Model* https://data.princeton.edu/wws509/notes/c7.pdf\n",
    "\n",
    "Trong trường hợp của Telco_Churn, ta có thể gán nhãn Death cho khách hàng rời bỏ dịch vụ, và Survival cho khách hàng vẫn còn duy trì dịch vụ. *Ý tưởng này được trình bày tại đây:* https://support.sas.com/resources/papers/proceedings/proceedings/sugi28/120-28.pdf . Từ ý tưởng này, sử dụng Telco_Churn dataset để trình bày diễn giải.\n",
    "\n",
    "*Đôi nét về phương pháp: Công thức KP bắt đầu từ năm 1952, khi ông còn giảng dạy ở Đại học Johns Hopkins. Lúc đó, Edward Kaplan đọc một bài báo của Greenwood (người nổi tiếng với công thức Greenwood trong phân tích ung thư) về thời gian bệnh nhân mắc bệnh ung thư. Kaplan liên tưởng đến cách ước tính tuổi thọ của linh kiện điện thoại chôn dưới đáy biển (mà sau này chúng ta biết đến dưới thuật ngữ survival analysis). Ông viết bài báo về cách phân tích survival và đưa bài đó cho John Tukey xem đồng thời nộp cho JASA để công bố. Tukey báo cho ông biết rằng Paul Meier cũng đang nghiên cứu về vấn đề này, và bản thảo cũng đã được nộp cho JASA. Sau này JASA đề nghị hai người (Kaplan và Meier) nên hợp tác viết chung một bài báo, thay vì đăng riêng 2 bài. Phải hơn 4 năm trao đổi và hoàn thiện, sau bài báo mới hoàn chỉnh và được JASA công bố vào năm 1958. Bài báo này khi mới công bố 3 năm đã có 1495 trích dẫn. Cho đến nay, bài báo đã được trích dẫn trên 35.000 lần, trở thành một trong những công trình có ảnh hưởng lớn nhất đến lịch sử khoa học!*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Công thức tổng quát của xác suất sống sót tại bất kỳ thời điểm cụ thể nào:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3059453/bin/IJAR-1-274-g001.jpg\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cách tính Ước tính Kaplan-Meier như sau:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3059453/bin/IJAR-1-274-g002.jpg\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Tương tự như vậy, gán nhãn `Death` cho khách hàng `Churn` (1) và `Survival` cho khách hàng `Not Churn` (0). Và cột `tenure` chính là thời gian tính từ lúc khách hàng đăng kí đến lúc khách hàng rời đi"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encoder cho feature `Churn`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "eda['Churn']=eda['Churn'].apply(lambda x: 1 if x == 'Yes' else 0 )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Quan sát các Kaplan-Meier và nhận xét"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import matplotlib.pyplot as plt\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1501846ac88>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Biểu đồ trên cho biết, sẽ có khoản 20% khách hàng rời đi vào ngày 20, 30% khách hàng rời đi vào ngày 50,...."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import matplotlib.pyplot as plt\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1501aa6c308>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Biểu đồ trên cho biết khách hàng sử dụng Subscribed Streaming TV có xác suất duy trì sử dụng dịch vụ lâu hơn so với khách hàng không đăng kí"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import matplotlib.pyplot as plt\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1501a962408>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Biểu đồ trên cho biết khách hàng sử dụng StreamingMovies có xác suất duy trì sử dụng dịch vụ lâu hơn so với khách hàng không đăng kí.\n",
    "\n",
    "Và các yếu tố khác cũng như vậy, biểu diễn từng feature để quan sát thêm. Mình pass qua vấn đề ở đây"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Mô hình Cox (Cox proportional hazards model)\n",
    "\n",
    "#### Đại ý của model: dự đoán xác xuất rời bỏ dịch vụ của một khách hàng cụ thể theo thời gian\n",
    "Thông thường nếu biến outcome là biến nhị phân (chỉ có hai giá trị 0/1, yes/no, sống/chết …) thì chúng ta dùng mô hình hồi qui logistic. Nhưng bài toán ở đây khác, biến outcome là biến nhị phân, mà nó lại phụ thuộc vào thời gian theo dõi. Ví dụ như bệnh nhân được điều trị, có người sống suốt thời gian theo dõi, có người chết sau điều trị vài tháng, có người sống đến 2 năm, lại có người chúng ta không biết còn sống hay chết vì không theo dõi được. Trong những trường hợp này, chúng ta dùng mô hình hồi quy Cox (còn gọi là mô hình hazards).\n",
    "\n",
    "Gọi L(t) là rủi ro (hazard) của một cá nhân tại thời điểm t, mô hình hồi qui Cox phát biểu rằng L(t) được xác định bởi tích số của rủi ro trung bình (L0) và ảnh hưởng của các yếu tố nguy cơ (X1, X2, …, Xk):\n",
    "\n",
    "L(t) = L0*exp(b1X1 + b2X2 + … + bkXk)\n",
    "\n",
    "Trong đó, b1, b2, …, bk là hệ số cần phải ước tính từ dữ liệu thực tế. Ví dụ như tôi muốn biết rủi ro bị ung thư trong vòng 5 năm, L(5), của một cá nhân bằng rủi ro trung bình trong quần thể, L0, và ảnh hưởng của các yếu tố như độ tuổi, giới tính, lối sống, gen, v.v."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Công thức tổng quát:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://miro.medium.com/max/1134/1*_HG7KpWGiGpT4gTVQWyxhg.png\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "from IPython.core.display import HTML \n",
    "Image(url= \"https://miro.medium.com/max/1134/1*_HG7KpWGiGpT4gTVQWyxhg.png\")    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Trực quan mô hình Cox và Model dự đoán thời xác xuất rời bỏ dịch vụ của một khách hàng cụ thể theo thời gian\n",
    "\n",
    "Ở bài viết này, mình dùng thư viện `CoxPHFitter` để xử lí `Survival regression`\n",
    "*đọc thêm về `CoxPHFitter` tại đây* :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from lifelines import CoxPHFitter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Sau khi đi qua Feture Selection, mình chọn top các feature sau để chạy mô hình"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import matplotlib.pyplot as plt\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import matplotlib.pyplot as plt\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <td>lifelines.CoxPHFitter</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>duration col</th>\n",
       "      <td>'tenure'</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>event col</th>\n",
       "      <td>'Churn'</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>baseline estimation</th>\n",
       "      <td>breslow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>number of observations</th>\n",
       "      <td>7042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>number of events observed</th>\n",
       "      <td>1869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>partial log-likelihood</th>\n",
       "      <td>-15182.29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time fit was run</th>\n",
       "      <td>2020-04-28 10:30:50 UTC</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div><table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coef</th>\n",
       "      <th>exp(coef)</th>\n",
       "      <th>se(coef)</th>\n",
       "      <th>coef lower 95%</th>\n",
       "      <th>coef upper 95%</th>\n",
       "      <th>exp(coef) lower 95%</th>\n",
       "      <th>exp(coef) upper 95%</th>\n",
       "      <th>z</th>\n",
       "      <th>p</th>\n",
       "      <th>-log2(p)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.99</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.99</td>\n",
       "      <td>0.99</td>\n",
       "      <td>-6.14</td>\n",
       "      <td>&lt;0.005</td>\n",
       "      <td>30.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <td>0.40</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.29</td>\n",
       "      <td>0.51</td>\n",
       "      <td>1.33</td>\n",
       "      <td>1.66</td>\n",
       "      <td>7.16</td>\n",
       "      <td>&lt;0.005</td>\n",
       "      <td>40.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gender_Male</th>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.99</td>\n",
       "      <td>0.05</td>\n",
       "      <td>-0.10</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.90</td>\n",
       "      <td>1.08</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Partner_Yes</th>\n",
       "      <td>-0.81</td>\n",
       "      <td>0.45</td>\n",
       "      <td>0.05</td>\n",
       "      <td>-0.92</td>\n",
       "      <td>-0.70</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.50</td>\n",
       "      <td>-14.93</td>\n",
       "      <td>&lt;0.005</td>\n",
       "      <td>165.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dependents_Yes</th>\n",
       "      <td>-0.36</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.07</td>\n",
       "      <td>-0.49</td>\n",
       "      <td>-0.22</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.80</td>\n",
       "      <td>-5.21</td>\n",
       "      <td>&lt;0.005</td>\n",
       "      <td>22.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PhoneService_Yes</th>\n",
       "      <td>0.69</td>\n",
       "      <td>2.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.49</td>\n",
       "      <td>0.90</td>\n",
       "      <td>1.63</td>\n",
       "      <td>2.45</td>\n",
       "      <td>6.66</td>\n",
       "      <td>&lt;0.005</td>\n",
       "      <td>35.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>StreamingTV_No internet service</th>\n",
       "      <td>-2.10</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.13</td>\n",
       "      <td>-2.36</td>\n",
       "      <td>-1.84</td>\n",
       "      <td>0.09</td>\n",
       "      <td>0.16</td>\n",
       "      <td>-15.80</td>\n",
       "      <td>&lt;0.005</td>\n",
       "      <td>184.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>StreamingTV_Yes</th>\n",
       "      <td>-0.19</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.06</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.07</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.93</td>\n",
       "      <td>-3.09</td>\n",
       "      <td>&lt;0.005</td>\n",
       "      <td>8.99</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table><div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Concordance</th>\n",
       "      <td>0.71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Log-likelihood ratio test</th>\n",
       "      <td>941.49 on 8 df</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>-log2(p) of ll-ratio test</th>\n",
       "      <td>655.08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_dummy = pd.get_dummies(df_r, drop_first=True)\n",
    "cph = CoxPHFitter()\n",
    "cph.fit(df_dummy, 'tenure', event_col='Churn')\n",
    "cph.print_summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import matplotlib.pyplot as plt\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1501a599d48>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (12,8))\n",
    "cph.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Tách thử khách hàng thứ 100 đến 110 để quan sát xác xuất rời bỏ dịch vụ theo thời gian"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import matplotlib.pyplot as plt\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1501a647108>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 864x576 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "tr_rows = df_dummy.iloc[100:110, 2:]\n",
    "plt.figure(figsize = (12,8))\n",
    "cph.predict_survival_function(tr_rows).plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Kết thúc bài viết ở đây, mong nhận nhiều ý kiến phản hồi của anh em"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}